Supporting optical fiber and component metrology needs requires development and evaluation of new measurement techniques, dissemination of this knowledge, and, when appropriate, development of Standard Reference Materials (SRM) or other calibration aids to help industry calibrate instrumentation. The project currently focuses on two areas: wavelength calibration standards in the near infrared region and nonlinear properties of optical fiber.

Wavelength standards are needed to calibrate instruments that measure the wavelengths of sources and characterize the wavelength dependence of components, such as those used in a WDM system. Fundamental references based on atomic and molecular absorption or emission lines provide the highest accuracy, but they are not available in all wavelength regions. The project currently produces four wavelength reference Standard Reference Materials (SRMs) based on fundamental molecular absorption lines: SRM 2514 (carbon monoxide ¹²C¹⁶O), SRM 2515 (carbon monoxide ¹³C¹⁶O), SRM 2517a (acetylene, high resolution), and SRM 2519 (hydrogen cyanide). Together these SRMs can be used to calibrate the wavelength scale of instruments between 1510 and 1630 nm.

WDM may expand into other wavelength regions, such as the 1300 to 1500 nm region. We have developed a hybrid multiple-wavelength reference that incorporates the wavelength flexibility of artifact references and the stability of fundamental molecular absorption references. Customized multiple wavelength reflectors can be generated by writing multiple superimposed fiber Bragg gratings (FBG) into optical fiber. Each grating is a reflector for a specific wavelength of light; the wavelengths are selected during the grating fabrication process. Strain and temperature changes can cause the center wavelength of these reflectors to change. On the other hand, atomic and molecular absorption lines are very stable under changing environmental conditions. If one of the FBG reflectors is located near an atomic or molecular absorption line, it can be actively stabilized to that line. This stability is then transferred to the other gratings, because they are superimposed at the same location in the fiber. We have demonstrated a hybrid reference based on this principle that provides multiple calibration references in the 850,1300, and 1550 nm regions.

Plans for 2004-2005:

• Release high-accuracy version of SRM 2519 (1530-1560 nm wavelength calibration)
  
• Deliver hybrid wavelength reference for the 850,1300, and 1550 nm regions to the Navy Primary Standards Laboratory

NIST needs higher accuracy internal wavelength references to calibrate its equipment. We have developed references for this purpose at 1560 nm and 1314 nm with uncertainties of a few megahertz. In collaboration with the NIST Time and Frequency Division, we have developed two accurate frequency combs in the telecommunication wavelength region to measure other references. To generate an optical frequency comb in the near infrared, pulses from either a Cr:forsterite femtosecond laser or a fiber laser are launched into highly nonlinear optical fiber, which broadens the laser spectrum to a supercontinuum with a width of > 1000 nm. Since the laser is pulsed, this broad supercontinuum will actually be composed of discrete frequency lines - a frequency comb - spaced by the laser repetition rate. This optical frequency comb can be stabilized and used to measure references throughout the 1100-2000 nm region. Any unknown optical frequency can be measured by simply comparing its frequency to that of the nearest tooth of the stabilized frequency comb. Using the Cr:forsterite laser-based frequency comb, we have measured several telecommunication optical frequency references, including three methane absorption lines in the 1300 nm region and a 1560 nm laser stabilized to a rubidium line. We have also developed a fiber laser-based frequency comb spanning from 1100 to 2200 nm that can be self-referenced to a microwave reference source.

• Develop fiber frequency comb into tool for frequency measurement
  
• Measure frequency of high accuracy reference candidates using frequency comb
  
• Develop robust, user-friendly near IR frequency comb for NIST internal use; transfer technology to industry

A striking example of the nonlinear effects in optical fiber is provided by the generation of extremely broad spectra in highly nonlinear fiber. These broad spectra, or supercontinua, can have widths spanning more than an octave in frequency and are generated by launching femtosecond pulses into specialty highly nonlinear optical fiber. Supercontinua can be generated in the visible using a Ti:sapphire laser or in the near infrared using either a fiber laser or a Cr:Forsterite laser. These supercontinua have a number of possible applications in telecommunications, including use as wavelength calibration references or as a WDM source after spectral slicing. In addition, these supercontinua will find uses in optical coherence tomography and spectroscopy. For many applications, the noise on the supercontinuum can be a limiting factor. Unfortunately, the same nonlinear processes that give rise to the supercontinuum, also amplify any input noise. In 2002-2003, we conducted a systematic study of the resulting amplitude noise across visible supercontinua generated using a Ti:sapphire laser. We identified both a low-frequency noise component that arises from the technical noise on the laser, and a broadband frequency noise component that arises from the initial shot noise on the input laser. In addition to this excess amplitude noise, we expect a high level of excess phase noise on supercontinua generated using a Ti:Sapphire laser. Moreover, this excess amplitude and phase noise is expected to be present in other supercontinuum systems as well, such as a fiber laser-based supercontinuum source, and will continue to be a limiting factor to applications of supercontinuum generation.

A second area of nonlinear effects that we have explored in some detail is Raman amplification. In Raman amplification, a strong pump beam amplifies a weaker signal beam through stimulated Raman scattering. In 2002-2003, we developed a simple technique for determining the full wavelength dependence of the Raman gain. Using this technique, we have participated in a TIA round robin measurement of Raman gain.

Plans for 2004-2005:

• Analyze the phase noise on visible supercontinuum generated from Ti:Sapphire laser light launched into microstructure fiber
  
• Explore phase and amplitude noise on fiber-laser based supercontinuum source
  
• Continue to support the TIA Raman gain round robin